Cleaning Buffer Preparation Tank Air-Liquid Interface Rings
Buffers are routinely used in biopharmaceutical manufacturing operations to adjust pH, salinity, or nutrient levels; equilibrate and flush columns and filter membranes; and for in-process and final product formulations. While buffers are easily cleaned with water (with a few exceptions due to high water solubility), the industry continues to struggle with visible residue at the air–liquid interface on buffer preparation and storage tanks. The residue can adhere tightly to the surface, appearing as distinct bands or rings at the air–liquid interface. Articles and presentations have been published that define the residue as hydrophobic in nature and due to trace polymers from the packaging, handling, raw materials, or manufacturing processes found in buffer components.1 ,2 ,3 These trace components, such as slip agents, are used to help the flow of dry ingredients, manipulation of plastic packaging, and storage of the components. Hydrophobic contaminants can also be derived from worn or chemically incompatible gaskets, valve diaphragms, and tubing materials. This article will review the different types of buffers, provide approaches to investigate the best cleaning procedure using laboratory testing, and detail strategies for cleaning air–liquid interface residues within buffer preparation and storage tanks.
Buffers are solutions whose pH is altered not to any great extent by the addition of small quantities of strong acid or strong base. Buffer solutions are divided into two types: 1. Acidic buffers are a mixture of a salt of a weak acid and a strong base, for example: acetic acid (CH3COOH) and sodium acetate (CH3COONa). A solution containing equal quantities of acetic acid and sodium acetate maintains a pH value around 4.75. pH = pKa + log (salt/acid) where Ka is the acid dissociation constant of the weak acid. 2. Basic buffers are a mixture of a salt of a weak base and a strong acid, such as ammonium hydroxide (NH4OH) and ammonium chloride (NH4Cl). A solution containing equal quantities of ammonium hydroxide and ammonium chloride maintains its pH value around 9.25. pOH = pKb + log (salt/base) where Kb is the base dissociation constant of the weak base. These equations are called Henderson–Hasselbalch equations. Buffers can be further divided:
- Single-substances solutions, which represent the solution of the salt of a weak acid and weak base.
- Mixture solutions, which represent any acid buffer or basic buffer.
- Natural buffer solutions, which maintain their pH in a variety of conditions. Human blood, for example maintains a pH value around 7.35 in spite of the wide variety of foods we consume.
Non-Routine Cleaning Challenges
Common challenges with cleaning buffer preparation tanks are the bands of residue found along air–liquid interfaces. These can occur after a single production batch or after multiple batches of the same or differing buffers. As reported in the 2013 CIP Summit9 benchmarking survey, 89% of the participating companies that use water for cleaning reported that their buffer preparation tanks were cleaned with purified water. Buffer salt solubility supports the use of “water-only” cleaning, but companies continue to be challenged with visible residue on the preparation tanks concentrated as bands or rings. Common approaches to cleaning air–liquid interface rings:2 ,6
- Increased spray impingement by using rotating spray devices focused on the air–liquid interface
- Increased time of the initial rinse, which typically removes gross residue
- Use of oxidizing agent with an alkaline cleaning solution
- Increased temperature (75–85°C)
- Use of a formulated cleaning agent containing surfactants and chelants
- Increased cleaning agent concentration
- Clean first with an acid, followed by an alkaline cleaning agent
Laboratory Testing Model
Laboratory testing has been effective at defining critical cleaning parameters for removing process residues.7 ,8 Laboratory testing involves applying the residue on a surface that is representative of the preparation and storage tank. The test continues by conditioning the applied residue on a surface to simulate the “real-world” process, and finally cleaning the surface in a manner that is representative of how the equipment is normally cleaned within the facility. The various cleaning factors to monitor during the laboratory testing include cleaning agent, temperature, time, action, water quality, surface material, rinsing method, and environmental factors. Purified water is commonly used as the cleaning agent to remove buffer residue between batches of the same buffer as well as different buffers. Water-only cleaning utilizes water solubility at the temperature cleaned as the cleaning mechanism. When a formulated cleaning agent is used, the cleaning mechanisms include solubility in an aqueous solution, wetting, emulsification, dispersion, chelation, and hydrolysis. These additional cleaning mechanisms are important in removing water-insoluble residue from the surface. Procedure and Results Some of the buffers evaluated include:
- Acetate, pH 5
- Acetate, pH 8
- Acetic acid, 1M
- Glucose buffer
- HEPES, 25mM
- Sodium acetate, 20mM
- Sodium chloride, 0.15M
- Sodium chloride, 3M
- Sodium chloride, 2.5M, pH 7
- Sodium chloride, 2M, pH 5.6
- Sodium chloride, 1M /acetic acid, 1M
- Sodium hydroxide, 0.1M
- Sodium hydroxide, 1M
- Sodium phosphate, 60mM
- TRIS buffer
- and more …
Table B outlines the laboratory test procedure.
A coupon was considered to be clean if it was visually clean, water break free, and its precoating and postcleaning weights were equal (< 0.1 mg residue per 7.5 × 15 cm coupon).8 Under these conditions, the buffers evaluated were easily cleaned in 5 minutes using deionized water at ambient temperature. Unfortunately, these cleaning parameters are not always effective in preventing visible residue in buffer preparation tanks.
One of the difficulties with providing effective cleaning recommendations for removing air–liquid interface rings is in generating laboratory coupons representative of what is seen on the tank walls. A number of factors can contribute to this:
- Trace elements are normally in parts per billion levels within the buffer solution.
- Trace elements are normally hydrophobic in nature and migrate to the air–liquid interface.
- High mixing speeds migrate the trace components to the air–liquid interface, which increases residue adherence to the tank side walls.
- Large liquid volume increases the amount of trace residues present.
- Small surface area at the air–liquid interface allows greater residue concentration.
- Repetitive hot water rinses can heat the surfaces and bake residues onto the side walls. Contaminants in the water can also have an impact. Characteristics of the hot water, such as its source before heating, should be considered.
In a mammalian cell culture performed by a large multinational biopharmaceutical company, for example, floats similar to fishing bobs were added to the bioreactor during production, and then sectioned for cleaning evaluation. This model was effective in identifying two distinct residue types at the air–liquid interface.
- A whitish residue more consistent with salts and antifoam in the media (Figure 1)
- A brownish residue consisting of proteins, lipids, and cellular debris used in the culture process (Figure 2)
Laboratory testing using agitated immersion proved that the whitish residue band required 4× the concentration of the formulated alkaline cleaning agent used to remove than the brown residue. Unfortunately, due to the high mixing rates used in media and buffer preparation tanks, 316L stainless steel floats are not a practical option because they can damage tank side walls. A different model is required—one that will generate a residue or change in the surface property of the coupon representative of the air–liquid interface residue along the side wall. Partially submerging the coupon into a beaker of media or buffer for a period of time will generate a visible residue at the air–liquid interface (Figure 3). But this residue is normally easy to clean, and may not be representative of the residue on the tank surface.
Residue conditioning techniques can increase the difficulty of cleaning the residue from the coupon, making it more representative of the actual interface residue in media and buffer tanks. Several conditioning techniques are listed below:
- Air dried at ambient temperature (Figure 4)
- Baked on the surface
- Coupon partially submerged and then air dried at ambient temperature
- Coupon partially submerged and then baked on the surface
- Coupon preheated, coated with the buffer, and then baked on the surface
Identifying the major component(s) in the air–liquid interface residue helps determine the cleaning parameters. This allows the analyst to spike the buffer solution with high levels of trace residue found in the buffer and at the air–liquid interface. Increasing soil levels can also decrease the time required to condition coupons and produce visible residue with tenacity similar to that observed in the production area. In our laboratory tests, we spiked buffer samples with slip agents to generate residue on stainless steel surfaces similar to what we have seen in the field. We tested several of the most common slip agents, wetting agents, and packaging materials used in the industry: erucamide, oleamide, stearamide, talc, polyisoprene, and polyethylene. Erucamide, oleamide, (Figure 5) and stearamide are the most commonly used slip agents in the pharmaceutical industry and life sciences industries. Because they are insoluble in water and buffers, they tend to float at the liquid surface, aggregate at the air–liquid interface, and adhere to the tank side wall.
Coupons were prepared under several different conditions:
- Slip agent mixed with the buffer
- Melted slip agent mixed with buffer
- Slip agent dissolved in methanol or ethanol and mixed with buffer
The coating process consisted of dispensing approximately 4 ml of samples onto the coupon with a plastic transfer pipette, and then spreading the sample over an approximately 100 cm2 area with the pipette. The amount of residue by weight and surface area coated were recorded to provide the amount of residue in mg/cm2 per sample. The hardest-to-clean condition was when the slip agent was dissolved in methanol or ethanol and then mixed with the buffer. Coated coupons were conditioned five ways:
- Air-dried at ambient temperature for 16 hours
- Baked in a drying oven at 121°C for 16 hours
- Autoclaved at 121°C for 1 hour
- Partially submerged in the cleaning solution and baked in a drying oven at 121°C for 16 hours
- Preheated in a drying oven at 121°C before coating and then baked in a drying oven at 121°C for 1 hour
Because condition 5 was hardest to clean, we chose it as our test model. We evaluated the following cleaning agents to determine the most suitable chemistry for removing erucamide, oelamide, and stearamide residues from the air–liquid interface (Table C):
- Deionized water
- 5% w/v sodium hydroxide or potassium hydroxide
- 5% v/v formulated alkaline cleaner containing sodium hydroxide or potassium hydroxide as well as surfactants and other components
- 2% v/v formulated alkaline cleaner containing potassium hydroxide + 2% v/v detergent additive containing hydrogen peroxide
Oxidative Chemistry Effectiveness
Cleaners formulated with sodium hypochlorite or hydrogen peroxide use oxidation as the cleaning mechanism. Oxidation cleaves high-molecular-weight molecules into smaller molecules, which are more susceptible to removal by other cleaning mechanisms, such as emulsion, solubility, and hydrolysis. Repeated use of oxidative cleaning agents at high concentration and temperature may discolor stainless steel, so these are generally used for periodic cleaning.
When assessing the effect of the air–liquid interface residue on the quality of the next product or batch, one should ask the questions shown in Table D.
Case Study 1
A pharmaceutical drug manufacturer was observing air–liquid interface “rings” in several blend tanks (Figure 6). The rings appeared at multiple levels in the tank and were black in color. The blended product was water soluble, so purified water was used to clean the tank between product batches.
After visual inspection of the tank, the air–liquid interface rings were wiped and scraped, and the residue was submitted for material analysis; a wipe sample of a black gasket was also submitted (Figure 7). The results of the scraped material were somewhat inconclusive as to its source, due to the opacity of the black particulates. However, Fourier transform infrared spectroscopy (FTIR) results from the gasket and tank ring wipes appeared similar.
Procedure The gasket, scraping material, and wipe samples of the gasket and tank ring were analyzed using a Digilab Excalibur Series FTIR FTS 4000 (SSR: 734) Fourier transform infrared spectrometer. Infrared spectroscopic data were collected between 4,000 cm-1 and 600 cm-1 (neat). A background spectrum was collected using the blank ATR diamond cell. Results The FTIR gasket spectra (carbon black filled) and the black scraping material produced skewed baselines. Carbon black filled types of materials are difficult to analyze by spectroscopic methods because of their very high absorption rates and their propensity to scatter infrared light. Figure 8 shows the spectrum of the gasket; Figure 9 shows the spectrum of the black scraping from the ring in the tank. Note that although the spectra are not exactly the same, there are similarities in some absorbance bands and the baseline slope.
Spectra of the wipe material (clean spot) and black residue on the wipe were collected independently. Spectral subtraction was used to determine the black residue on the wiping material. The spectra from gasket and tank ring wipes (Figure 10) appear to be fairly similar after subtraction of the wipe material.
The tank rings contained residue from the black EPDM gaskets. Water-only cleaning was not sufficient to remove the hydrophobic residue, which allowed it to build up on the tank as visible residue. The solution was filtered into a storage tank where there were no observed rings. Further testing should be done to identify the residue and assess its effect on the product. Gaskets were replaced and the tanks were cleaned with a combination of manual cleaning (brush from the manway) and automated cleaning using a spray device with a formulated alkaline cleaning agent containing surfactants. The tanks and gaskets were also monitored more closely to determine if the gaskets should be replaced more frequently.
Case Study 2
A biopharmaceutical manufacturer was observing air–liquid interface rings in its sodium chloride buffer tanks. The sodium chloride arrives as large clumps in polyethylene containers with a polyethylene bag liner (Figure 11). The clumps are broken up by the operators while in the container prior to preparing a sodium chloride solution. The buffer tanks are cleaned with water between batches. Laboratory studies to investigate residue ring cleaning and prevention were conducted in parallel with field trials and analytical residue testing. Procedure Laboratory testing consisted of two parts:
- The 304 stainless steel coupons were partially submerged in a beaker with 3M sodium chloride solution or a saturated sodium chloride solution for 30 minutes while mixing on a stir plate. The soiled coupons were then air dried at ambient temperature overnight (> 16 hours) prior to washing with 80°C deionized water by agitated immersion for 5 minutes. The soiling and cleaning process was repeated up to 10 times. The clean coupons were visually inspected, and then rinsed with tap water for 10 seconds at 2 gallons per minute per foot bandwidth. The coupons were rinsed on each side with de-ionized water and examined for a water break-free surface. The coupons were then air dried at ambient temperature and visually observed for cleanliness. Gravimetric analysis and FTIR analysis was also performed on select samples.
- Part 2 testing compared repeated soiling and cleaning as detailed above, but coupons were washed with a 0.25% v/v formulated alkaline CIP detergent at 80°C for 5 minutes by agitated immersion instead of 80°C deionized water by agitated immersion for 5 minutes.
Results While the coupons were visually clean, those cleaned only with water failed the water break free test after only one soiling and cleaning cycle (Figure 12). The water break free failure was not produced on stainless steel when cleaned with 0.25% v/v formulated alkaline CIP detergent by agitated immersion at 80°C for 5 minutes, even after 10 soiling and cleaning cycles. Analytical testing of the air–liquid interface rings demonstrated trace polyethylene residue, most likely from the raw material packaging. The results did not confirm conclusively that the residue observed in the laboratory studies was polyethylene, however. These laboratory studies demonstrated that the sodium chloride contained a water-insoluble hydrophobic residue that remained on the surface when cleaned only with water. This residue could continue to build up on the surface and result in a visual failure or reduce the efficiency of subsequent cleaning, sanitization, or stainless steel maintenance. A low-concentration alkaline cleaning agent with surfactants appeared to remove this hydrophobic residue when used after each cleaning. Periodic cleaning of the surface with a low-concentration alkaline cleaning agent after the residue develops to a visual failure was not evaluated during this study.
This article reviewed common buffers used in pharmaceutical and biopharmaceutical manufacturing processes. Based on solubility and benchtop cleaning trials, water should be effective at removing theses residue. However, trace components from the packaging, raw materials, gaskets, diaphragms, tubing, etc., can migrate to the tank surface during blending and adhere to the side walls, creating air–liquid interface rings after one or more batches. Identifying the ring components is the first step in determining whether the residue is intrinsic (e.g., residue from a minor component in a raw material, such as starch or a wetting agent in calcium carbonate) or extrinsic (e.g., residue from packaging material or a worn gasket) to the buffer. Laboratory studies have been effective at providing a course of action in some cases, but simulating the air–liquid interface residue without having identified the residue remains a challenge. Identifying the residue allows analysts to spike the buffer sample, which improves the visibility and tenaciousness of the residue on test coupons. Cleaning procedures could also be modified to better remove extrinsic particulates, or at least implement a maintenance cleaning procedure proven to clean the equipment and prevent the ring from forming or remove it once it is visible. If the residue is inherent to the process, the cleaning procedure should be modified to consistently clean the buffer residue; at minimum, a maintenance cleaning procedure should be defined to clean the equipment prior to the residue being visible. An alternative would be to identify the source of trace material and try to eliminate it from the raw material through a corrective and preventive action plan. If the air–liquid interface ring is due to a maintenance issue, then the source of the residue should be identified, the equipment repaired or replaced, and cleaned. Common cleaning approaches include the use of a formulated alkaline cleaning agent at elevated temperatures, often with a detergent additive to increase the surfactant level with or without hydrogen peroxide. If this is not successful then a manual cleaning of the air–liquid interface may be warranted. By: Paul Lopolito, Dijana Hadziselimovic, Amanda Deal, and Amy Thanavaro
About the Authors
Paul Lopolito is a technical services manager for the Life Sciences Division of STERIS Corporation (Mentor, Ohio, US). He currently provides global technical support related to process cleaning and contamination control, which includes field support, site audits, training presentations and educational seminars. Paul has more than 15 years of industry experience and has held positions as a technical services manager, manufacturing manager and laboratory manager. He has authored and published numerous articles on cleaning and contamination control. He earned a B.A. in Biological Sciences from Goucher College in Towson, Maryland, US. Dijana Hadziselimovic is a senior technical services associate for the Life Sciences Division of STERIS Corporation (Mentor, Ohio, US). She provides technical support in the area of process and research detergents and conducts laboratory experiments to recommend cleaning procedures. Dijana has over 13 years of laboratory experience in the pharmaceutical and biotech industries. She holds a BA degree in chemistry from the University of Missouri, St. Louis (Missouri, US). Amanda Deal is a senior technical services associate for the Life Sciences Division of STERIS Corporation (Mentor, Ohio). She provides technical support in the area of process research detergents and conducts laboratory experiments to recommend cleaning procedures. Amanda has over nine years of laboratory experience in the pharmaceutical and biotech industries. She holds a Bachelor’s of Science degree in Biology from the University of Missouri–Kansas City. Amy Thanavaro, PhD is a Senior Scientist, Analytical Services & Development, STERIS Corporation. Her current responsibilities include scientific leadership of new product development and developing analytical analyses to support Product Development, Manufacturing and cross-site functional projects. She led the chemistry team and coordinated with multiple teams that resulted in the 510(k) Clearance for the Reliance Endoscope Processing System (Cleaning Claim). Dr. Thanavaro received her B.S. degree in chemistry from Chulalongkorn University, Bangkok, Thailand, in 1996 and her Ph.D. in chemistry from University of Missouri–St. Louis in 2003.
- 1Sharnez, R., C. Flint, M. Monk, and A. Tholudur. “Cleaning Validation Challenges for Bioprocesses Strategies for Eliminating Tenacious Residues.” Presented at the PDA Annual Meeting, Phoenix, Arizona, US, 16–18 April 2012.
- 2 a b Lopolito, P. “Case Study: Cleaning and Passivation of Buffer Preparation Tanks.” Presented at the Cleaning Validation and Critical Cleaning Processes Conference, Philadelphia, Pennsylvania, US, 19–21 August 2013.
- 3Carlson, J. (2014) “Design a Cleaning and Passivation Program for Buffer Preparation Tanks.” Presented at the Cleaning Validation and Critical Cleaning Processes Conference, San Diego, CA, 20 August 2014.
- 4Merck. The Merck Index, 12th ed. CRC Press: Whitehouse Station, New Jersey, US, 1996.
- 5Stevens, L. “Buffers,” chapter 3 in Proteins Lab Fax, edited by N.C. Price. Academic Press, 1996.
- 9 2013 CIP Technology and Industry Benchmarking Summit. Seventh Annual CIP Technology & Industry Benchmarking Summit. Benchmarking Sessions at Amgen in Longmont, Colorado, US, 23 April 2013.
- 6Carlson, J. “Biotech Residues and Cleaning,” chapter 17 in Cleaning and Cleaning Validation, Volume 1, edited by P.L. Pluta. Parenteral Drug Association and Davis Healthcare International Publishing, 2009, pp. 383–413.
- 7Rathore, N., W. Qi, C. Chen, and W. Ji. “Bench-Scale Characterization of Cleaning Process Design Space for Biopharmaceuticals.” Biopharm International 22, no. 3 (1 March 2009): 32–44.
- 8 a b Hadziselimovic, D., and P Lopolito.” Critical Cleaning of Carbomer Containing Products.” Journal of GXP Compliance 16, no. 3 (2012): 32–38.